U.S. patent application number 16/734252 was filed with the patent office on 2021-07-08 for out of order wave slot release for a terminated wave.
The applicant listed for this patent is QUALCOMM INCORPORATED. Invention is credited to Yun DU, Andrew Evan GRUBER, Baoguang YANG, Zilin YING, Chun YU.
Application Number | 20210209717 16/734252 |
Document ID | / |
Family ID | 1000004611919 |
Filed Date | 2021-07-08 |
United States Patent
Application |
20210209717 |
Kind Code |
A1 |
DU; Yun ; et al. |
July 8, 2021 |
OUT OF ORDER WAVE SLOT RELEASE FOR A TERMINATED WAVE
Abstract
Methods, systems, and devices for image processing are
described. A device may determine, based on a test operation, to
terminate a first wave associated with a first slot of a set of
slots. The device may update a terminated wave bit associated with
the first slot based on the determination to terminate the first
wave. In some aspects, the device may update a number of
invocations field associated with the first wave based on the
determination to terminate the first wave. The device may release
the first slot based on updating the terminated wave bit and the
number of invocations field. In some examples, the device may
output the number of invocations field to a rendering backend of
the device based on the terminated wave bit.
Inventors: |
DU; Yun; (San Diego, CA)
; YU; Chun; (Rancho Santa Fe, CA) ; GRUBER; Andrew
Evan; (Arlington, MA) ; YING; Zilin; (San
Diego, CA) ; YANG; Baoguang; (Fremont, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
QUALCOMM INCORPORATED |
San Diego |
CA |
US |
|
|
Family ID: |
1000004611919 |
Appl. No.: |
16/734252 |
Filed: |
January 3, 2020 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G06T 1/20 20130101; G06T
11/001 20130101 |
International
Class: |
G06T 1/20 20060101
G06T001/20; G06T 11/00 20060101 G06T011/00 |
Claims
1. A method for image processing at a device, comprising:
determining, based at least in part on a test operation, to
terminate a first wave associated with a first slot of a set of
slots; updating a terminated wave bit associated with the first
slot based at least in part on the determination to terminate the
first wave; updating a number of invocations field associated with
the first wave based at least in part on the determination to
terminate the first wave; releasing the first slot based at least
in part on updating the terminated wave bit and the number of
invocations field; and outputting the number of invocations field
to a rendering backend of the device based at least in part on the
terminated wave bit.
2. The method of claim 1, further comprising: performing the test
operation for the first wave; identifying a number of pixels
associated with the first wave based at least in part on the test
operation; and determining, based at least in part on the test
operation, to terminate all pixels of the identified number of
pixels, wherein the determination to terminate the first wave is
based at least in part on the determination to terminate all pixels
of the identified number of pixels.
3. The method of claim 2, wherein updating the number of
invocations field associated with the first wave comprises:
updating the number of invocations field associated with the first
wave to indicate the identified number of pixels.
4. The method of claim 1, further comprising: identifying a wave
order associated with a set of waves based at least in part on a
sequence queue, wherein the set of waves comprises at least the
first wave; and identifying the terminated wave bit based at least
in part on the identified wave order, wherein the number of
invocations field is output based at least in part on the
identified terminated wave bit and the identified wave order.
5. The method of claim 1, wherein updating the terminated wave bit
and the number of invocations field comprises: writing the
terminated wave bit and the number of invocations field to an entry
of a sequence queue corresponding to the first wave.
6. The method of claim 5, further comprising: identifying the
terminated wave bit based at least in part on the sequence queue,
wherein the number of invocations field is output based at least in
part on the identified terminated wave bit and the identified wave
order.
7. The method of claim 1, further comprising: receiving, at the
rendering backend, the output number of invocations field; and
identifying a pixel ordering for rendering based at least in part
on the received number of invocations field.
8. The method of claim 1, further comprising: processing one or
more pixels of a second wave using a second slot; processing one or
more pixels of a third wave using the released first slot; and
processing one or more pixels of a fourth wave using a third
slot.
9. The method of claim 8, further comprising: writing the
terminated wave bit and the number of invocations field to a first
entry of a sequence queue corresponding to the first wave; writing
the one or more pixels of the second wave to a second entry of the
sequence queue corresponding to the second wave; writing the one or
more pixels of the third wave to a third entry of the sequence
queue corresponding to the third wave; and writing the one or more
pixels of the fourth wave to a fourth entry of the sequence queue
corresponding to the fourth wave.
10. The method of claim 9, further comprising: outputting, by a
streaming processor of the device, the first entry, the second
entry, the third entry, and the fourth entry to the rendering
backend of the device based at least in part on the sequence queue;
and identifying a pixel ordering for rendering pixel information
based at least in part on the number of invocations field
associated with the first wave.
11. The method of claim 8, further comprising: performing a second
test operation for the second wave; and identifying a number of
pixels associated with the second wave based at least in part on
the second test operation, wherein the one or more pixels comprise
a subset of the number of pixels associated with the second
wave.
12. The method of claim 11, further comprising: outputting color
information associated with the one or more pixels to the rendering
backend of the device based at least in part on processing the one
or more pixels of the second wave.
13. An apparatus for image processing at a device, comprising: a
processor, memory coupled with the processor; and instructions
stored in the memory and executable by the processor to cause the
apparatus to: determine, based at least in part on a test
operation, to terminate a first wave associated with a first slot
of a set of slots; update a terminated wave bit associated with the
first slot based at least in part on the determination to terminate
the first wave; update a number of invocations field associated
with the first wave based at least in part on the determination to
terminate the first wave; release the first slot based at least in
part on updating the terminated wave bit and the number of
invocations field; and output the number of invocations field to a
rendering backend of the device based at least in part on the
terminated wave bit.
14. The apparatus of claim 13, wherein the instructions are further
executable by the processor to cause the apparatus to: perform the
test operation for the first wave; identify a number of pixels
associated with the first wave based at least in part on the test
operation; and determine, based at least in part on the test
operation, to terminate all pixels of the identified number of
pixels, wherein the determination to terminate the first wave is
based at least in part on the determination to terminate all pixels
of the identified number of pixels.
15. The apparatus of claim 14, wherein the instructions to update
the number of invocations field associated with the first wave are
executable by the processor to cause the apparatus to: update the
number of invocations field associated with the first wave to
indicate the identified number of pixels.
16. The apparatus of claim 13, wherein the instructions are further
executable by the processor to cause the apparatus to: identify a
wave order associated with a set of waves based at least in part on
a sequence queue, wherein the set of waves comprises at least the
first wave; and identify the terminated wave bit based at least in
part on the identified wave order, wherein the number of
invocations field is output based at least in part on the
identified terminated wave bit and the identified wave order.
17. The apparatus of claim 13, wherein the instructions to update
the terminated wave bit and the number of invocations field are
executable by the processor to cause the apparatus to: write the
terminated wave bit and the number of invocations field to an entry
of a sequence queue corresponding to the first wave.
18. The apparatus of claim 17, wherein the instructions are further
executable by the processor to cause the apparatus to: identify the
terminated wave bit based at least in part on the sequence queue,
wherein the number of invocations field is output based at least in
part on the identified terminated wave bit and the identified wave
order.
19. The apparatus of claim 13, wherein the instructions are further
executable by the processor to cause the apparatus to: receive, at
the rendering backend, the output number of invocations field; and
identify a pixel ordering for rendering based at least in part on
the received number of invocations field.
20. An apparatus for image processing at a device, comprising:
means for determining, based at least in part on a test operation,
to terminate a first wave associated with a first slot of a set of
slots; means for updating a terminated wave bit associated with the
first slot based at least in part on the determination to terminate
the first wave; means for updating a number of invocations field
associated with the first wave based at least in part on the
determination to terminate the first wave; means for releasing the
first slot based at least in part on updating the terminated wave
bit and the number of invocations field; and means for outputting
the number of invocations field to a rendering backend of the
device based at least in part on the terminated wave bit.
Description
BACKGROUND
[0001] The following relates generally to image processing, and
more specifically to out of order wave slot release for a
terminated wave.
[0002] Multimedia systems are widely deployed to provide various
types of multimedia communication content such as voice, video,
packet data, messaging, broadcast, and so on. These multimedia
systems may be capable of processing, storage, generation,
manipulation and rendition of multimedia information. Examples of
multimedia systems include entertainment systems, information
systems, virtual reality systems, model and simulation systems, and
so on. These systems may employ a combination of hardware and
software technologies to support processing, storage, generation,
manipulation and rendition of multimedia information, for example,
such as capture devices, storage devices, communication networks,
computer systems, and display devices.
[0003] For example, a graphics processing unit (GPU) may represent
one or more dedicated processors for performing graphical
operations. A GPU may be a dedicated hardware unit having fixed
function and programmable components for rendering graphics and
executing GPU applications. For example, a GPU may include a
plurality of processing elements, which may allow the GPU to
generate graphic images for display (e.g., for graphical user
interfaces, for display of two-dimensional or three-dimensional
graphics scenes, etc.).
SUMMARY
[0004] The described techniques relate to improved methods,
systems, devices, and apparatuses that support out of order wave
slot release for a terminated wave. For example, as discussed
herein, graphics processing units (GPUs) may be parallel
processors, capable of processing large amounts of data at the same
time. In some cases, a GPU may implement a lot of threads (e.g., or
pixels in pixel shader implementations) running the same shader. As
such, threads (e.g., or pixels, workloads, etc.) may be batched in
groups, which may be referred to as waves (e.g., or wavefronts,
warps, etc.). A number of threads in a wave may be architecture
dependent (e.g., architectures may implement 32 threads in a wave,
64 threads in a wave, etc.). All the threads that are in a wave may
run the same shader in lockstep. As such, a wave may generally
refer to a smallest unit of scheduled work for a compute unit (CU)
(e.g., a Graphics Core Next (GCN) CU) to run (e.g., where each wave
may include, for example, 64 threads).
[0005] In some cases, a GPU may terminate (e.g., suspend, kill,
etc.) a wave, however processing resources (e.g., a wave slot
associated with the terminated wave) may remain idle in order to
wait for local memory operations to complete. For example, a next
wave may not start until older waves complete as to not disrupt
output ordering from GPU local memory to downstream blocks (e.g.,
such as a rendering backend performing post color processing on
wave output), such that a wave slot associated with a terminated
wave may remain idle (e.g., which may be inefficient in terms of
wave slot resource utilization). The techniques described herein
may provide for out of order wave slot release for a terminated
wave, while maintaining output order (e.g., to a rendering
backend).
[0006] A method of image processing at a device is described. The
method may include determining, based on a test operation, to
terminate a first wave associated with a first slot of a set of
slots, updating a terminated wave bit associated with the first
slot based on the determination to terminate the first wave, and
updating a number of invocations field associated with the first
wave based on the determination to terminate the first wave. The
method may further include releasing the first slot based on
updating the terminated wave bit and the number of invocations
field, and outputting the number of invocations field to a
rendering backend of the device based on the terminated wave
bit.
[0007] An apparatus for image processing at a device is described.
The apparatus may include a processor, memory coupled with the
processor, and instructions stored in the memory. The instructions
may be executable by the processor to cause the apparatus to
determine, based on a test operation, to terminate a first wave
associated with a first slot of a set of slots, update a terminated
wave bit associated with the first slot based on the determination
to terminate the first wave, and update a number of invocations
field associated with the first wave based on the determination to
terminate the first wave. The instructions may be executable by the
processor to further cause the apparatus to release the first slot
based on updating the terminated wave bit and the number of
invocations field, and output the number of invocations field to a
rendering backend of the device based on the terminated wave
bit.
[0008] Another apparatus for image processing at a device is
described. The apparatus may include means for determining, based
on a test operation, to terminate a first wave associated with a
first slot of a set of slots, updating a terminated wave bit
associated with the first slot based on the determination to
terminate the first wave, and updating a number of invocations
field associated with the first wave based on the determination to
terminate the first wave. The apparatus may further include means
for releasing the first slot based on updating the terminated wave
bit and the number of invocations field, and outputting the number
of invocations field to a rendering backend of the device based on
the terminated wave bit.
[0009] A non-transitory computer-readable medium storing code for
image processing at a device is described. The code may include
instructions executable by a processor to determine, based on a
test operation, to terminate a first wave associated with a first
slot of a set of slots, update a terminated wave bit associated
with the first slot based on the determination to terminate the
first wave, and update a number of invocations field associated
with the first wave based on the determination to terminate the
first wave. The code may include instructions further executable by
a processor to release the first slot based on updating the
terminated wave bit and the number of invocations field, and output
the number of invocations field to a rendering backend of the
device based on the terminated wave bit.
[0010] Some examples of the method, apparatuses, and non-transitory
computer-readable medium described herein may further include
operations, features, means, or instructions for performing the
test operation for the first wave, identifying a number of pixels
associated with the first wave based on the test operation, and
determining, based on the test operation, to terminate all pixels
of the identified number of pixels, where the determination to
terminate the first wave may be based on the determination to
terminate all pixels of the identified number of pixels. In some
examples of the method, apparatuses, and non-transitory
computer-readable medium described herein, updating the number of
invocations field associated with the first wave may include
operations, features, means, or instructions for updating the
number of invocations field associated with the first wave to
indicate the identified number of pixels.
[0011] Some examples of the method, apparatuses, and non-transitory
computer-readable medium described herein may further include
operations, features, means, or instructions for identifying a wave
order associated with a set of waves based on a sequence queue,
where the set of waves includes at least the first wave, and
identifying the terminated wave bit based on the identified wave
order, where the number of invocations field may be output based on
the identified terminated wave bit and the identified wave
order.
[0012] In some examples of the method, apparatuses, and
non-transitory computer-readable medium described herein, updating
the terminated wave bit and the number of invocations field may
include operations, features, means, or instructions for writing
the terminated wave bit and the number of invocations field to an
entry of a sequence queue corresponding to the first wave. Some
examples of the method, apparatuses, and non-transitory
computer-readable medium described herein may further include
operations, features, means, or instructions for identifying the
terminated wave bit based on the sequence queue, where the number
of invocations field may be output based on the identified
terminated wave bit and the identified wave order. Some examples of
the method, apparatuses, and non-transitory computer-readable
medium described herein may further include operations, features,
means, or instructions for receiving, at the rendering backend, the
output number of invocations field, and identifying a pixel
ordering for rendering based on the received number of invocations
field.
[0013] Some examples of the method, apparatuses, and non-transitory
computer-readable medium described herein may further include
operations, features, means, or instructions for processing one or
more pixels of a second wave using a second slot, processing one or
more pixels of a third wave using the released first slot, and
processing one or more pixels of a fourth wave using a third slot.
Some examples of the method, apparatuses, and non-transitory
computer-readable medium described herein may further include
operations, features, means, or instructions for writing the
terminated wave bit and the number of invocations field to a first
entry of a sequence queue corresponding to the first wave, writing
the one or more pixels of the second wave to a second entry of the
sequence queue corresponding to the second wave, writing the one or
more pixels of the third wave to a third entry of the sequence
queue corresponding to the third wave, and writing the one or more
pixels of the fourth wave to a fourth entry of the sequence queue
corresponding to the fourth wave.
[0014] Some examples of the method, apparatuses, and non-transitory
computer-readable medium described herein may further include
operations, features, means, or instructions for outputting, by a
streaming processor of the device, the first entry, the second
entry, the third entry, and the fourth entry to the rendering
backend of the device based on the sequence queue, and identifying
a pixel ordering for rendering pixel information based on the
number of invocations field associated with the first wave. Some
examples of the method, apparatuses, and non-transitory
computer-readable medium described herein may further include
operations, features, means, or instructions for performing a
second test operation for the second wave, and identifying a number
of pixels associated with the second wave based on the second test
operation, where the one or more pixels include a subset of the
number of pixels associated with the second wave. Some examples of
the method, apparatuses, and non-transitory computer-readable
medium described herein may further include operations, features,
means, or instructions for outputting color information associated
with the one or more pixels to the rendering backend of the device
based on processing the one or more pixels of the second wave.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] FIG. 1 illustrates an example of a system for image
processing that supports out of order wave slot release for a
terminated wave in accordance with aspects of the present
disclosure.
[0016] FIG. 2 illustrates an example of a device that supports out
of order wave slot release for a terminated wave in accordance with
aspects of the present disclosure.
[0017] FIG. 3 illustrates an example of a processing diagram that
supports out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure.
[0018] FIGS. 4A and 4B illustrate example processing diagrams that
support out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure.
[0019] FIGS. 5 and 6 show block diagrams of devices that support
out of order wave slot release for a terminated wave in accordance
with aspects of the present disclosure.
[0020] FIG. 7 shows a block diagram of a GPU that supports out of
order wave slot release for a terminated wave in accordance with
aspects of the present disclosure.
[0021] FIG. 8 shows a diagram of a system including a device that
supports out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure.
[0022] FIGS. 9 and 10 show flowcharts illustrating methods that
support out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure.
DETAILED DESCRIPTION
[0023] A graphics processing unit (GPU) may represent one or more
dedicated processors for performing graphical operations. A GPU may
be a dedicated hardware unit having fixed function and programmable
components for rendering graphics and executing GPU applications.
In some cases, a GPU may implement a parallel processing structure
that may provide for more efficient processing of complex
graphic-related operations. For example, a GPU may include a
plurality of processing elements that are configured to operate in
a parallel manner, which may allow the GPU to generate graphic
images for display (e.g., for graphical user interfaces, for
display of two-dimensional or three-dimensional graphics scenes,
etc.).
[0024] In some cases, a GPU may execute a shader (e.g., a shader
program, a graphics program, a computer program, etc.) to perform
various specialized functions in various fields of computer
graphics. For example, shaders may be used widely in cinema
postprocessing, computer-generated imagery, multimedia display,
video games, etc. (e.g., to produce a wide range of effects such as
production of appropriate levels of light, darkness, and color
within an image, such as alteration of hue, saturation, brightness
or contrast of an image, such as production of blur, light bloom,
volumetric lighting, normal mapping for depth effects, distortion,
such as edge detection, motion detection, and many others). A GPU
may thus execute a shader to transform two-dimensional or
three-dimensional data into useful two-dimensional data for
displaying.
[0025] Shaders may be written to apply transformations to a large
set of elements at a time (e.g., to each pixel in an area of a
screen, or for every vertex of a model). GPUs may thus execute
shaders via parallel processing, and a GPU may have multiple shader
pipelines (e.g., slots, threads, etc.) to facilitate such parallel
processing and improve computation throughput. Efficient compute
operations may often rely on getting waves synchronized in a
desired pattern (e.g., in a desired slot pattern). GPUs may
implement systems of wave scheduling, however in some cases such
scheduling may interfere with or contradict execution for some
types of workloads. In some cases, GPUs may manage or process
non-uniform and irregular workloads for graphics such that in
implementation synchronization patterns may be less than ideal
(e.g., some waves may wait for other waves to catch up or finish
processing, which may result in processing delays, rendering
latency, etc.).
[0026] For example, in some cases, streaming processors (SPs) may
process input waves (e.g., pixel warps, workloads, work items,
etc.) via multiple parallel wave slots according to an input order
(priority order) associated with the waves. As the waves are
processed, the SP may dispatch (e.g., output) results to a
rendering backend (RB) in the order the waves are input to the
parallel wave slots. However, processing inefficiencies may arise
when a wave being processed in a wave slot is terminated (e.g.,
killed). For example, when a wave (e.g., a Pixel_warp W1) being
processed in an available wave slot (S1) is killed, SPs may not
immediately reallocate the wave slot (S1) for another wave to be
processed. Instead, the wave slot may remain occupied (e.g., even
though the wave is terminated such that the slot may be idle) until
all higher priority waves (Pixel_warp W0) have been processed
according to the input (priority) order (e.g., in order to preserve
output or dispatch order to a rendering backend).
[0027] According to the techniques described herein, an event queue
(e.g., a sequence queue) of a SP may include a terminated wave bit
(e.g., for indicating whether a wave has been terminated) and a
number of invocations field (e.g., a quad count number, a pixel
count number, a fragments number, etc. for indicating the number of
killed invocations, pixels, quads, etc.). When a wave is
terminated, the SP may update the terminated wave bit and the
number of invocations field, and the SP may release the wave slot
for incoming workloads (e.g., other waves to be processed). Based
on the updated terminated wave bit, a dispatcher may output the
number of killed invocations to the RB. Accordingly, the SP may
output processed waves (e.g., pixel information, color information,
etc. associated with the processed waves) to the RB without
transferring color information of killed pixels (e.g., of killed
waves) while maintaining output order to the RB.
[0028] For example, a device (e.g., a GPU) may run or perform a
test operation (e.g., a visibility pass, lighting test, shadow
test, etc.) to determine whether or not to terminate (kill) one or
more pixels of a wave (e.g., to determine whether one or more
pixels of a wave will be visible after rendering). In some cases,
after performing such a test, the device may determine to terminate
all pixels of the wave and thus terminate the wave. Some workloads
(e.g., waves) may be processed by a SP dynamically, such that the
oldest workloads may be associated with higher priority to be
processed and output as soon as possible. Therefore, because some
devices (e.g., some GPUs) may use internal memory for storing pixel
locations information, color information, etc., such information
stored in internal memory may be dispatched downstream (e.g., to a
RB) in order for post color processing (e.g., as the downstream
block to assume the dispatched or output pixel information is in
order based on workload/wave priority). In some aspects, the test
operation may be included in a shader (e.g., a shader instruction).
For example, the device (e.g., the GPU) may process one or more
shaders (e.g., shader instructions), where the shaders may include
operations (e.g., test operations) for determining whether or not
to terminate (kill) one or more pixels of a wave.
[0029] According to the techniques described herein, the SP (e.g.,
the dispatcher) may inform a downstream block of the number of
pixels killed (e.g., via a terminated wave bit and a number of
invocations field) such that the downstream block may know which
pixels are terminated to maintain expected ordering (e.g., for post
color operations). As such, a slot associated with a terminated
wave may be released to free up the slot resources for processing
of a subsequent wave (e.g., rather than the slot remaining idle to
preserve downstream output/dispatch ordering). The described
techniques may provide for resource return and resource reuse in an
efficient manner with little additional overhead, which may improve
GPU performance (e.g., improve GPU resource utilization, reduce
parallel slot processing interruptions, improve wave
synchronization, provide greater control for non-uniform or
irregular workloads, etc.).
[0030] Aspects of the disclosure are initially described in the
context of a multimedia system. Aspects of the disclosure are
further illustrated by and described with reference to apparatus
diagrams, system diagrams, and flowcharts that relate to out of
order wave slot release for a terminated wave.
[0031] FIG. 1 illustrates a multimedia system 100 for a device that
supports out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure. The multimedia
system 100 may include devices 105, a server 110, and a database
115. Although, the multimedia system 100 illustrates two devices
105, a single server 110, a single database 115, and a single
network 120, the present disclosure applies to any multimedia
system architecture having one or more devices 105, servers 110,
databases 115, and networks 120. The devices 105, the server 110,
and the database 115 may communicate with each other and exchange
information that supports out of order wave slot release for a
terminated wave, such as multimedia packets, multimedia data, or
multimedia control information, via network 120 using
communications links 125. In some cases, a portion or all of the
techniques described herein supporting out of order wave slot
release for a terminated wave may be performed by the devices 105
or the server 110, or both.
[0032] A device 105 may be a cellular phone, a smartphone, a
personal digital assistant (PDA), a wireless communication device,
a handheld device, a tablet computer, a laptop computer, a cordless
phone, a display device (e.g., monitors), and/or the like that
supports various types of communication and functional features
related to multimedia (e.g., transmitting, receiving, broadcasting,
streaming, sinking, capturing, storing, and recording multimedia
data). A device 105 may, additionally or alternatively, be referred
to by those skilled in the art as a user equipment (UE), a user
device, a smartphone, a Bluetooth device, a Wi-Fi device, a mobile
station, a subscriber station, a mobile unit, a subscriber unit, a
wireless unit, a remote unit, a mobile device, a wireless device, a
wireless communications device, a remote device, an access
terminal, a mobile terminal, a wireless terminal, a remote
terminal, a handset, a user agent, a mobile client, a client,
and/or some other suitable terminology. In some cases, the devices
105 may also be able to communicate directly with another device
(e.g., using a peer-to-peer (P2P) or device-to-device (D2D)
protocol). For example, a device 105 may be able to receive from or
transmit to another device 105 variety of information, such as
instructions or commands (e.g., multimedia-related
information).
[0033] The devices 105 may include an application 130 and a
multimedia manager 135. While, the multimedia system 100
illustrates the devices 105 including both the application 130 and
the multimedia manager 135, the application 130 and the multimedia
manager 135 may be an optional feature for the devices 105. In some
cases, the application 130 may be a multimedia-based application
that can receive (e.g., download, stream, broadcast) from the
server 110, database 115 or another device 105, or transmit (e.g.,
upload) multimedia data to the server 110, the database 115, or to
another device 105 via using communications links 125.
[0034] The multimedia manager 135 may be part of a general-purpose
processor, a digital signal processor (DSP), an image signal
processor (ISP), a central processing unit (CPU), a graphics
processing unit (GPU), a microcontroller, an application-specific
integrated circuit (ASIC), a field-programmable gate array (FPGA),
a discrete gate or transistor logic component, a discrete hardware
component, or any combination thereof, or other programmable logic
device, discrete gate or transistor logic, discrete hardware
components, or any combination thereof designed to perform the
functions described in the present disclosure, and/or the like. For
example, the multimedia manager 135 may process multimedia (e.g.,
image data, video data, audio data) from and/or write multimedia
data to a local memory of the device 105 or to the database
115.
[0035] The multimedia manager 135 may also be configured to provide
multimedia enhancements, multimedia restoration, multimedia
analysis, multimedia compression, multimedia streaming, and
multimedia synthesis, among other functionality. For example, the
multimedia manager 135 may perform white balancing, cropping,
scaling (e.g., multimedia compression), adjusting a resolution,
multimedia stitching, color processing, multimedia filtering,
spatial multimedia filtering, artifact removal, frame rate
adjustments, multimedia encoding, multimedia decoding, and
multimedia filtering. By further example, the multimedia manager
135 may process multimedia data to support out of order wave slot
release for a terminated wave, according to the techniques
described herein.
[0036] The server 110 may be a data server, a cloud server, a
server associated with an multimedia subscription provider, proxy
server, web server, application server, communications server, home
server, mobile server, or any combination thereof. The server 110
may in some cases include a multimedia distribution platform 140.
The multimedia distribution platform 140 may allow the devices 105
to discover, browse, share, and download multimedia via network 120
using communications links 125, and therefore provide a digital
distribution of the multimedia from the multimedia distribution
platform 140. As such, a digital distribution may be a form of
delivering media content such as audio, video, images, without the
use of physical media but over online delivery mediums, such as the
Internet. For example, the devices 105 may upload or download
multimedia-related applications for streaming, downloading,
uploading, processing, enhancing, etc. multimedia (e.g., images,
audio, video). The server 110 may also transmit to the devices 105
a variety of information, such as instructions or commands (e.g.,
multimedia-related information) to download multimedia-related
applications on the device 105.
[0037] The database 115 may store a variety of information, such as
instructions or commands (e.g., multimedia-related information).
For example, the database 115 may store multimedia 145. The device
may support out of order wave slot release for a terminated wave
associated with the multimedia 145. The device 105 may retrieve the
stored data from the database 115 via the network 120 using
communication links 125. In some examples, the database 115 may be
a relational database (e.g., a relational database management
system (RDBMS) or a Structured Query Language (SQL) database), a
non-relational database, a network database, an object-oriented
database, or other type of database, that stores the variety of
information, such as instructions or commands (e.g.,
multimedia-related information).
[0038] The network 120 may provide encryption, access
authorization, tracking, Internet Protocol (IP) connectivity, and
other access, computation, modification, and/or functions. Examples
of network 120 may include any combination of cloud networks, local
area networks (LAN), wide area networks (WAN), virtual private
networks (VPN), wireless networks (using 802.11, for example),
cellular networks (using third generation (3G), fourth generation
(4G), long-term evolved (LTE), or new radio (NR) systems (e.g.,
fifth generation (5G)), etc. Network 120 may include the
Internet.
[0039] The communications links 125 shown in the multimedia system
100 may include uplink transmissions from the device 105 to the
server 110 and the database 115, and/or downlink transmissions,
from the server 110 and the database 115 to the device 105. The
wireless communication links 125 may transmit bidirectional
communications and/or unidirectional communications. In some
examples, the communication links 125 may be a wired connection or
a wireless connection, or both. For example, the communications
links 125 may include one or more connections, including but not
limited to, Wi-Fi, Bluetooth, Bluetooth low-energy (BLE), cellular,
Z-WAVE, 802.11, peer-to-peer, LAN, wireless local area network
(WLAN), Ethernet, FireWire, fiber optic, and/or other connection
types related to wireless communication systems.
[0040] The device 105 may determine, based on a test operation, to
terminate a first wave associated with a first slot of a set of
slots. In some aspects, the device 105 may update a terminated wave
bit associated with the first slot based on the determination to
terminate the first wave. In some examples, the device 105 may
update a number of invocations field associated with the first wave
based on the determination to terminate the first wave. The device
105 may release the first slot based on updating the terminated
wave bit and the number of invocations field. In an example, the
device 105 may output the number of invocations field to a
rendering backend of the device based on the terminated wave
bit.
[0041] The techniques described herein may provide improvements in
identifying wave slots associated with terminated (e.g., killed)
waves, as well as improvements in resource (e.g., slot) utilization
(e.g., via improved wave slot release for incoming waves,
workloads, etc.). Furthermore, the techniques described herein may
provide enhancements to the operation of devices 105. For example,
by releasing a slot associated with a terminated wave to free up
the slot resources for processing of a subsequent wave (e.g.,
rather than the slot remaining idle to preserve downstream
output/dispatch ordering), the operational characteristics, such as
power consumption, processor utilization (e.g., DSP, CPU, GPU, ISP
processing utilization), and memory usage of the devices 105 may be
reduced. The techniques described herein may also provide improved
efficiency associated with resource return and resource reuse of
the devices 105 (e.g., resource return and resource reuse
associated with wave slots of a streaming processor), with little
additional overhead, by reducing latency associated with processes
related to out of order wave slot release for a terminated
wave.
[0042] FIG. 2 illustrates an example of a device 200 in accordance
with various aspects of the present disclosure. In some cases,
device 200 may implement aspects of out of order wave slot release
for a terminated wave performed by a device 105 as described with
reference to FIG. 1. Examples of device 200 include, but are not
limited to, wireless devices, mobile or cellular telephones,
including smartphones, personal digital assistants (PDAs), video
gaming consoles that include video displays, mobile video gaming
devices, mobile video conferencing units, laptop computers, desktop
computers, televisions set-top boxes, tablet computing devices,
e-book readers, fixed or mobile media players, and the like.
[0043] In the example of FIG. 2, device 200 includes a central
processing unit (CPU) 210 having CPU memory 215, a GPU 225 having
GPU memory 230 and an SP 250, a display 245, a display buffer 235
storing data associated with rendering, a user interface unit 205,
and a system memory 240. For example, system memory 240 may store a
GPU driver 220 (illustrated as being contained within CPU 210 as
described herein) having a compiler, a GPU program, a
locally-compiled GPU program, and the like. User interface unit
205, CPU 210, GPU 225, system memory 240, and display 245 may
communicate with each other (e.g., using a system bus).
[0044] Examples of CPU 210 include, but are not limited to, a
digital signal processor (DSP), general purpose microprocessor,
ASIC, FPGA, or other equivalent integrated or discrete logic
circuitry. Although CPU 210 and GPU 225 are illustrated as separate
units in the example of FIG. 2, in some examples, CPU 210 and GPU
225 may be integrated into a single unit. CPU 210 may execute one
or more software applications. Examples of the applications may
include operating systems, word processors, web browsers, e-mail
applications, spreadsheets, video games, audio and/or video
capture, playback or editing applications, or other such
applications that initiate the generation of image data to be
presented via display 245. As illustrated, CPU 210 may include CPU
memory 215. For example, CPU memory 215 may represent on-chip
storage or memory used in executing machine or object code. CPU
memory 215 may include one or more volatile or non-volatile
memories or storage devices, such as flash memory, a magnetic data
media, an optical storage media, etc. CPU 210 may be able to read
values from or write values to CPU memory 215 more quickly than
reading values from or writing values to system memory 240, which
may be accessed, e.g., over a system bus.
[0045] GPU 225 may represent one or more dedicated processors for
performing graphical operations. That is, for example, GPU 225 may
be a dedicated hardware unit having fixed function and programmable
components for rendering graphics and executing GPU applications.
GPU 225 may also include a DSP, a general purpose microprocessor,
an ASIC, an FPGA, or other equivalent integrated or discrete logic
circuitry. GPU 225 may be built with a highly-parallel structure
that provides more efficient processing of complex graphic-related
operations than CPU 210. For example, GPU 225 may include a
plurality of processing elements that are configured to operate on
multiple vertices or pixels in a parallel manner. The highly
parallel nature of GPU 225 may allow GPU 225 to generate graphic
images (e.g., graphical user interfaces and two-dimensional or
three-dimensional graphics scenes) for display 245 more quickly
than CPU 210.
[0046] GPU 225 may, in some instances, be integrated into a
motherboard of device 200. In other instances, GPU 225 may be
present on a graphics card that is installed in a port in the
motherboard of device 200 or may be otherwise incorporated within a
peripheral device configured to interoperate with device 200. As
illustrated, GPU 225 may include GPU memory 230. For example, GPU
memory 230 may represent on-chip storage or memory used in
executing machine or object code. GPU memory 230 may include one or
more volatile or non-volatile memories or storage devices, such as
flash memory, a magnetic data media, an optical storage media, etc.
GPU 225 may be able to read values from or write values to GPU
memory 230 more quickly than reading values from or writing values
to system memory 240, which may be accessed, e.g., over a system
bus. That is, GPU 225 may read data from and write data to GPU
memory 230 without using the system bus to access off-chip memory.
This operation may allow GPU 225 to operate in a more efficient
manner by reducing the need for GPU 225 to read and write data via
the system bus, which may experience heavy bus traffic.
[0047] As discussed, GPU 225 may represent one or more dedicated
processors for performing graphical operations. GPU 225 may be a
dedicated hardware unit having fixed function and programmable
components for rendering graphics and executing GPU applications.
In some cases, GPU 225 may implement a parallel processing
structure that may provide for more efficient processing of complex
graphic-related operations. For example, GPU 225 may include a
plurality of processing elements (e.g., which in some cases may
refer to slots, etc.) that may be configured to operate in a
parallel manner, which may allow the GPU 225 to generate graphic
images for display (e.g., for graphical user interfaces, for
display of two-dimensional or three-dimensional graphics scenes,
etc.).
[0048] As illustrated, GPU 225 may include SP 250. SP 250 may be
configured to process workloads associated with one or more
operations. Each workload, for example, may include multiple
threads, and multiple threads may be grouped, for example, based on
code (e.g., instructions) associated with the threads. In some
aspects, SP 250 may execute the multiple threads or multiple
threadgroups in parallel. SP 250 may be configured to dynamically
process the workloads. In some examples, SP 250 may prioritize
processing of the workloads (e.g., prioritize data associated with
the workloads) so as to improve resource utilization within the SP
250. In some examples, SP 250 may determine to terminate a wave
associated with a slot, and may release the slot (e.g., while
maintaining output order to a RB of the device 200) according to
the techniques described herein (e.g., according to implementation
of a terminated wave bit and a number of invocations field via a
sequence queue, etc.).
[0049] Display 245 represents a unit capable of displaying video,
images, text or any other type of data for consumption by a viewer.
Display 245 may include a liquid-crystal display (LCD), a light
emitting diode (LED) display, an organic LED (OLED), an
active-matrix OLED (AMOLED), or the like. Display buffer 235
represents a memory or storage device dedicated to storing data for
presentation of imagery, such as computer-generated graphics, still
images, video frames, or the like for display 245. Display buffer
235 may represent a two-dimensional buffer that includes a
plurality of storage locations. The number of storage locations
within display buffer 235 may, in some cases, generally correspond
to the number of pixels to be displayed on display 245. For
example, if display 245 is configured to include 640.times.480
pixels, display buffer 235 may include 640.times.480 storage
locations storing pixel color and intensity information, such as
red, green, and blue pixel values, or other color values. Display
buffer 235 may store the final pixel values for each of the pixels
processed by GPU 225. Display 245 may retrieve the final pixel
values from display buffer 235 and display the final image based on
the pixel values stored in display buffer 235.
[0050] User interface unit 205 represents a unit with which a user
may interact with or otherwise interface to communicate with other
units of device 200, such as CPU 210. Examples of user interface
unit 205 include, but are not limited to, a trackball, a mouse, a
keyboard, and other types of input devices. User interface unit 205
may also be, or include, a touch screen and the touch screen may be
incorporated as part of display 245.
[0051] System memory 240 may comprise one or more computer-readable
storage media. Examples of system memory 240 include, but are not
limited to, a random access memory (RAM), static RAM (SRAM),
dynamic RAM (DRAM), a read-only memory (ROM), an electrically
erasable programmable read-only memory (EEPROM), a compact disc
read-only memory (CD-ROM) or other optical disc storage, magnetic
disc storage, or other magnetic storage devices, flash memory, or
any other medium that can be used to store desired program code in
the form of instructions or data structures and that can be
accessed by a computer or a processor. System memory 240 may store
program modules and/or instructions that are accessible for
execution by CPU 210. Additionally, system memory 240 may store
user applications and application surface data associated with the
applications. System memory 240 may in some cases store information
for use by and/or information generated by other components of
device 200. For example, system memory 240 may act as a device
memory for GPU 225 and may store data to be operated on by GPU 225
as well as data resulting from operations performed by GPU 225
[0052] In some examples, system memory 240 may include instructions
that cause CPU 210 or GPU 225 to perform the functions ascribed to
CPU 210 or GPU 225 in aspects of the present disclosure. System
memory 240 may, in some examples, be considered as a non-transitory
storage medium. The term "non-transitory" should not be interpreted
to mean that system memory 240 is non-movable. As one example,
system memory 240 may be removed from device 200 and moved to
another device. As another example, a system memory substantially
similar to system memory 240 may be inserted into device 200. In
certain examples, a non-transitory storage medium may store data
that can, over time, change (e.g., in RAM).
[0053] System memory 240 may store a GPU driver 220 and compiler, a
GPU program, and a locally-compiled GPU program. The GPU driver 220
may represent a computer program or executable code that provides
an interface to access GPU 225. CPU 210 may execute the GPU driver
220 or portions thereof to interface with GPU 225 and, for this
reason, GPU driver 220 is shown in the example of FIG. 2 within CPU
210. GPU driver 220 may be accessible to programs or other
executables executed by CPU 210, including the GPU program stored
in system memory 240. Thus, when one of the software applications
executing on CPU 210 requires graphics processing, CPU 210 may
provide graphics commands and graphics data to GPU 225 for
rendering to display 245 (e.g., via GPU driver 220).
[0054] In some cases, the GPU program may include code written in a
high level (HL) programming language, e.g., using an application
programming interface (API). Examples of APIs include Open Graphics
Library ("OpenGL"), DirectX, Render-Man, WebGL, or any other public
or proprietary standard graphics API. The instructions may also
conform to so-called heterogeneous computing libraries, such as
Open-Computing Language ("OpenCL"), DirectCompute, etc. In general,
an API includes a predetermined, standardized set of commands that
are executed by associated hardware. API commands allow a user to
instruct hardware components of a GPU 225 to execute commands
without user knowledge as to the specifics of the hardware
components. In order to process the graphics rendering
instructions, CPU 210 may issue one or more rendering commands to
GPU 225 (e.g., through GPU driver 220) to cause GPU 225 to perform
some or all of the rendering of the graphics data. In some
examples, the graphics data to be rendered may include a list of
graphics primitives (e.g., points, lines, triangles,
quadrilaterals, etc.).
[0055] The GPU program stored in system memory 240 may invoke or
otherwise include one or more functions provided by GPU driver 220.
CPU 210 generally executes the program in which the GPU program is
embedded and, upon encountering the GPU program, passes the GPU
program to GPU driver 220. CPU 210 executes GPU driver 220 in this
context to process the GPU program. That is, for example, GPU
driver 220 may process the GPU program by compiling the GPU program
into object or machine code executable by GPU 225. This object code
may be referred to as a locally-compiled GPU program. In some
examples, a compiler associated with GPU driver 220 may operate in
real-time or near-real-time to compile the GPU program during the
execution of the program in which the GPU program is embedded. For
example, the compiler generally represents a unit that reduces HL
instructions defined in accordance with a HL programming language
to low-level (LL) instructions of a LL programming language. After
compilation, these LL instructions are capable of being executed by
specific types of processors or other types of hardware, such as
FPGAs, ASICs, and the like (including, but not limited to, CPU 210
and GPU 225).
[0056] In the example of FIG. 2, the compiler may receive the GPU
program from CPU 210 when executing HL code that includes the GPU
program. That is, a software application being executed by CPU 210
may invoke GPU driver 220 (e.g., via a graphics API) to issue one
or more commands to GPU 225 for rendering one or more graphics
primitives into displayable graphics images. The compiler may
compile the GPU program to generate the locally-compiled GPU
program that conforms to a LL programming language. The compiler
may then output the locally-compiled GPU program that includes the
LL instructions. In some examples, the LL instructions may be
provided to GPU 225 in the form a list of drawing primitives (e.g.,
triangles, rectangles, etc.).
[0057] The LL instructions (e.g., which may alternatively be
referred to as primitive definitions) may include vertex
specifications that specify one or more vertices associated with
the primitives to be rendered. The vertex specifications may
include positional coordinates for each vertex and, in some
instances, other attributes associated with the vertex, such as
color coordinates, normal vectors, and texture coordinates. The
primitive definitions may include primitive type information,
scaling information, rotation information, and the like. Based on
the instructions issued by the software application (e.g., the
program in which the GPU program is embedded), GPU driver 220 may
formulate one or more commands that specify one or more operations
for GPU 225 to perform in order to render the primitive. When GPU
225 receives a command from CPU 210, it may decode the command and
configure one or more processing elements to perform the specified
operation and may output the rendered data to display buffer
235.
[0058] GPU 225 generally receives the locally-compiled GPU program,
and then, in some instances, GPU 225 renders one or more images and
outputs the rendered images to display buffer 235. For example, GPU
225 may generate a number of primitives to be displayed at display
245. Primitives may include one or more of a line (including
curves, splines, etc.), a point, a circle, an ellipse, a polygon
(e.g., a triangle), or any other two-dimensional primitive. The
term "primitive" may also refer to three-dimensional primitives,
such as cubes, cylinders, sphere, cone, pyramid, torus, or the
like. Generally, the term "primitive" refers to any basic geometric
shape or element capable of being rendered by GPU 225 for display
as an image (or frame in the context of video data) via display
245. GPU 225 may transform primitives and other attributes (e.g.,
that define a color, texture, lighting, camera configuration, or
other aspect) of the primitives into a so-called "world space" by
applying one or more model transforms (which may also be specified
in the state data). Once transformed, GPU 225 may apply a view
transform for the active camera (which again may also be specified
in the state data defining the camera) to transform the coordinates
of the primitives and lights into the camera or eye space. GPU 225
may also perform vertex shading to render the appearance of the
primitives in view of any active lights. GPU 225 may perform vertex
shading in one or more of the above model, world, or view
space.
[0059] Once the primitives are shaded, GPU 225 may perform
projections to project the image into a canonical view volume.
After transforming the model from the eye space to the canonical
view volume, GPU 225 may perform clipping to remove any primitives
that do not at least partially reside within the canonical view
volume. That is, GPU 225 may remove any primitives that are not
within the frame of the camera. GPU 225 may then map the
coordinates of the primitives from the view volume to the screen
space, effectively reducing the three-dimensional coordinates of
the primitives to the two-dimensional coordinates of the screen.
Given the transformed and projected vertices defining the
primitives with their associated shading data, GPU 225 may then
rasterize the primitives. Generally, rasterization may refer to the
task of taking an image described in a vector graphics format and
converting it to a raster image (e.g., a pixelated image) for
output on a video display or for storage in a bitmap file
format.
[0060] A GPU 225 may include a dedicated fast bin buffer (e.g., a
fast memory buffer, such as global memory (GMEM), which may be
referred to by GPU memory 230). As discussed herein, a rendering
surface may be divided into bins. In some cases, the bin size is
determined by format (e.g., pixel color and depth information) and
render target resolution divided by the total amount of GMEM. The
number of bins may vary based on device 200 hardware, target
resolution size, and target display format. A rendering pass may
draw (e.g., render, write, etc.) pixels into GMEM (e.g., with a
high bandwidth that matches the capabilities of the GPU). The GPU
225 may then resolve the GMEM (e.g., burst write blended pixel
values from the GMEM, as a single layer, to a display buffer 235 or
a frame buffer in system memory 240). Such may be referred to as
bin-based or tile-based rendering. When all bins are complete, the
driver may swap buffers and start the binning process again for a
next frame.
[0061] For example, GPU 225 may implement a tile-based architecture
that renders an image or rendering target by breaking the image
into multiple portions, referred to as tiles or bins. The bins may
be sized based on the size of GPU memory 230 (e.g., which may
alternatively be referred to herein as GMEM or a cache), the
resolution of display 245, the color or Z precision of the render
target, etc. When implementing tile-based rendering, GPU 225 may
perform a binning pass and one or more rendering passes. For
example, with respect to the binning pass, GPU 225 may process an
entire image and sort rasterized primitives into bins.
[0062] FIG. 3 illustrates an example GPU 300 that supports out of
order wave slot release for a terminated wave in accordance with
aspects of the present disclosure. In some examples, the example
GPU 300 may implement aspects of multimedia system 100. GPU 300,
for example, may implement aspects of devices 105 and device 200.
GPU 300 may include an SP 305, a high level sequencer (HLSQ) 335, a
test block 340, UCHE 350 (e.g., unified L2 cache), rendering
backend (RB) 355, and a primitive processor 345.
[0063] The SP 305 may include aspects of streaming processors as
described herein. For example, SP 305 may be an example of or refer
to of aspects of SP 250, streaming processor 630, and streaming
processor 720, as described herein. In some aspects, SP 305 may
prioritize processing of workloads (e.g., prioritize data
associated with the workloads) so as to improve resource
utilization within the SP 305. The SP 305 may include a sequencer
315, local memory 320, a wave scheduler 310 (e.g., a wave scheduler
and context register), a TEX/load controller 330 (e.g., texel/load
controller), and a distributor 325 (e.g., a dispatcher). SP 305 may
include one or more physical processors located on GPU 300
configured to run or execute fragment and vertex shaders.
[0064] According to the techniques described herein, GPU 300 may
support out of order resource (e.g., wave slot) release for
terminated waves. For example, SP 305 may implement a terminated
wave bit and a number of invocations field, such that the RB 355
may identify a terminated wave and the number of pixels associated
with the terminated wave (e.g., such that slot resources associated
with the terminated wave may be reallocated while maintaining
ordering information at RB 355). For instance, GPU 300 may use
internal memory (e.g., local memory 320) for storing pixel
information (e.g., pixel location information, pixel color
information, etc.). As such, SP 305 may wait (e.g., based on local
memory 320 usage to reduce UCHE 350 cost) for such pixel
information to be dispatched (e.g., via distributor 325) downstream
(e.g., to RB 355) for post color processing. As such, a downstream
block (e.g., RB 355) may interpret color information from
distributor 325 according to some order.
[0065] When all pixels of a wave are terminated or killed, the
described techniques may take advantage of not needing to output
any data or color information for any of the pixels of the
terminated wave (e.g., such that the wave slot associated with the
terminated wave may be released quickly and made available for
subsequent waves). The wave slot may be released, and SP 305 (e.g.,
via sequencer 315, distributor 325, etc.) may generate, sequence
and output a terminated wave bit and a number of invocations field,
such that the RB 355 may identify a wave was killed (e.g., based on
the terminated wave bit) and how many pixels or quads are
associated with the killed wave (e.g., based on the number of
invocations field). Further, the RB 355 may thus maintain pixel
ordering (e.g., ordering of color information output from
distributor 325) without requiring additional tracking logic. As
such, reduced output (e.g., reduced overhead) associated with the
described techniques (e.g., due to no downstream pixel information
associated with cases where all pixels are terminated of a
terminated wave) may provide for resource release (e.g., wave slot
release) in an efficient manner.
[0066] In some cases, in addition to graphics operations (e.g.,
such as triangle rasterization to generate pixel information, pixel
processing discussed herein, etc.), SP 305 may further perform
compute operations. For example, compute operations may form work
item based execution for computations. In some cases, compute
operations may not be associated with an ordering requirement
(e.g., an eldest scheme) as outputs may go to system memory.
Generally, the techniques described herein may additionally apply
to compute operations (e.g., slots may be released for next
workloads), as well as other operations, by analogy, without
departing from the scope of the present disclosure. Test block 340
may generally perform early rejection of objects of any kind (e.g.,
objects, draw calls, tiles, triangles, pixels, etc.) that don't
contribute to the final image.
[0067] In some aspects, test block 340 may perform early rejection
of objects via one or more test operations, culling operations,
rasterization operations, etc. which may reject, for example,
triangles, different tiles, etc. In some aspects, test block 340
may perform early rejection of objects, for example, during
rasterization (e.g., tiles may be dropped, primitives that are
below a display size threshold may be rejected). For example, test
block 340 may perform a test operation (e.g., a visibility pass, a
culling operation, etc.), may determine to terminate invocations
(e.g., pixels, quads, etc.) of a wave, may determine to terminate a
wave, etc. In some cases, test block 340 may exchange (e.g., pass)
information (e.g., pixel coefficient information, culling
information, termination information, etc.) to local memory 320, to
distributor 325, etc. In some cases, some or all aspects of test
block 340 may refer to or be implemented by SP 305. TEX/load
controller 330 may store and load information (e.g., such as pixel
coefficient information) from local memory 320.
[0068] In some aspects, SP 305 may include one or more streaming
processors configured to process input waves (e.g., pixel warps,
workloads, work items, etc.) via multiple parallel wave slots. SP
305 may be configured to process the input waves according to an
input order (priority order) associated with the waves. As the
waves are processed, the SP 305 (e.g., via distributor 325) may
dispatch (e.g., output) results to a RB 355 based on a wave order
in which the waves are input to the parallel wave slots. For
example, workloads (e.g., waves) may be processed in SP 305
dynamically. In some cases, the oldest workloads may have priority
such that the oldest workloads may be processed as soon as
possible, via available slot resources, such that the oldest
workloads may be output to the RB 355 as soon as possible. In some
cases, RB 355 may access or utilize memory via UCHE 350.
[0069] In some cases, sequencer 315 of SP 305 may send information
(e.g., thread information) to wave scheduler 310, may send
information (e.g., attribute information) to TEX/Load controller
330, etc. In some aspects, sequencer 315 may allocate wave slots
and associated graphics processing space based on pixel tile
workload (e.g., i/j--barycentric coefficient data) issued to SP
305, for example, by HLSQ 335. For example, in some cases,
sequencer 315 may receive vertex thread (Vtx thread) information,
vertex attribute (Vtx Attribute) information, pixel thread
information, pixel attribute information, etc. from HLSQ 335. In an
example, sequencer 315 may then send thread information to wave
scheduler 310 and attribute information to TEX/Load controller 330
(e.g., in accordance with aspects of sequencing described
herein).
[0070] In some example aspects, the wave scheduler 310 of SP 305
may output an indication of terminated (e.g., killed) waves. In
some aspects, the wave scheduler 310 may output a terminated wave
bit (e.g., a toggle bit corresponding to the terminated wave, an
integer, a set of integers, etc.) indicative of the terminated
wave. In some examples, the SP 305 (e.g., the wave scheduler 310)
may output a number of invocations field, which may include a quad
number (e.g., number of quads, number of pixels, number of
fragments, etc.) associated with the terminated wave. In some
examples, the SP 305 (e.g., the wave scheduler 310) may terminate
waves and output the indication of the terminated waves based on a
test operation (e.g., a visibility pass, lighting test, shadow
test, etc.) performed by a device 105.
[0071] The wave scheduler 310 of SP 305 may manage ordering (e.g.,
reordering) associated with processing pixel shaders (e.g., the
wave scheduler 310 may manage wave scheduling), based on wave kill
commands issued by the device 105 (e.g., terminate wave or kill
wave commands issued by the GPU 300) with respect to a test
operation. For example, wave scheduler 310 may execute one or more
pixel shaders (e.g., schedule processing of pixel shaders and
allocate or reallocate wave slots associated with processing the
pixel shaders). In cases where a wave is terminated (e.g., in cases
where all pixels of a wave are killed and the wave is terminated),
wave scheduler 310 may reallocate resources (e.g., slots)
associated with the terminated wave as described in more detail
herein.
[0072] In some aspects, wave scheduler 310 may also be referred to
as a thread scheduler. The wave scheduler 310 may assign one or
more threads or threadgroups according to resource allocations,
processes, hardware, etc. associated with the threads or
threadgroups. In some aspects, wave scheduler 310 may store
information indicating the assignments, for example, in one or more
context registers associated with each thread.
[0073] In some aspects, wave scheduler 310 may exchange
instructions associated with processing the pixel shaders (e.g.,
processing the waves) with TEX/load controller 330, etc. In an
example, wave scheduler 310 may execute or implement processing
instructions (e.g., addresses, predicate, etc.) associated with
pixel shader processing (e.g., via elementary function units
(EFUs), full arithmetic logic units (FALUs), half arithmetic logic
units (HALUs), etc.). SP 305 may refrain from executing or
processing waves which are indicated by the wave scheduler 310 as
terminated (e.g., killed) waves.
[0074] According to examples of aspects of the techniques described
herein, when a wave being processed in a wave slot is terminated
(e.g., killed), SP 305 may improve processing efficiency through
wave slot allocation (e.g., reallocation) at wave scheduler 310.
For example, when a wave W1 (also referred to as a Pixel_warp W1)
being processed in a wave slot S1 is terminated, SP 305 may improve
processing efficiency through allocation (e.g., reallocation) of
wave slot S1 by wave scheduler 310.
[0075] In some aspects, wave scheduler 310 may reallocate the wave
slot for processing another wave to be processed (e.g., wave
scheduler 310 may reallocate the wave slot S1 for processing a wave
W4 to be processed). In an example, among a set of waves W1 through
W4 to be processed, wave W4 may have a lower priority compared to
waves W0, W2, and W3. In an example aspect, wave scheduler 310 may
reallocate the wave slot W1 for processing W4, such that wave W4
may be processed at wave slot W1 prior to the completed execution
of waves W0, W2, and W3, which (e.g., according to the input order)
have a higher priority than wave W4. For example, SP 305 may
process wave W4 without waiting until processing of waves W0, W2,
or W3 is complete. Based on the reallocation, idle time of wave
slots associated with a killed wave may be reduced or eliminated
(e.g., idle time of the wave slot S1 may be reduced or
eliminated).
[0076] Aspects of SP 305 may be advantageous over some SPs. For
example, for some SPs, when a wave being processed in a wave slot
is terminated (e.g., killed), processing by the SPs may be
inefficient. For example, when a wave W1 (also referred to as a
Pixel_warp W1) being processed in a wave slot S1 is terminated,
some SPs may not reallocate the wave slot S1 for processing another
wave to be processed (e.g., some SPs may not immediately reallocate
the wave slot S1 for a wave to be processed). That is, the wave
slot may remain occupied until all higher priority waves (e.g.,
according to the input order) have been processed. For example, for
a wave W4 to be processed, some SPs may wait to reallocate the wave
slot S1 allocated for wave W1 until waves W0, W2, W3 having higher
priority compared to the wave W4 have been processed, even when
wave W1 has already been terminated (e.g., as further described
herein, for example, with reference to FIG. 4A).
[0077] Based on the wave slot processing (e.g., the pixel shader
processing) according to aspects described herein, for example,
based on completion of the wave slot processing, SP 305 may
dispatch processed results to a downstream block such as RB 355. In
some aspects, the processed results may include color information.
SP 305 may, for example, dispatch the processed results to the RB
355 via an output interface such as distributor 325. Distributor
325 may include a demultiplexer, and in some examples, may be
referred to as a dispatcher. In some aspects, SP 305 (e.g., wave
scheduler 310) may direct TEX/load controller 330 to read-out
buffered results (e.g., from a general purpose register (GPR)) to
the distributor 325.
[0078] The distributor 325 (e.g., which in some cases may be
referred to as a dispatcher) may output pixels associated with
waves (e.g., output pixels associated with non-killed waves) to RB
355. For example, the distributor 325 may output color information
of pixels associated with non-killed waves to RB 355. In some
aspects, the distributor 325 may refrain from outputting pixels
(e.g., color information of pixels) associated with killed waves to
RB 355. In some example aspects, for a terminated (e.g., killed)
wave, the distributor 325 may output the terminated wave bit, the
number of invocations field (e.g., quads) associated with the
terminated wave, or a combination thereof to RB 355.
[0079] In some aspects, the distributor 325 may output vertex
specifications to primitive processor 345. The vertex
specifications may be included in, for example, LL instructions
(primitive definitions) as described herein. For example, the
vertex specifications may specify one or more vertices associated
with primitives to be rendered.
[0080] In some aspects of the techniques described, SP 305 may
manage ordering associated with processing a set of waves, which
may reduce overhead at the rendering backend (e.g., may reduce
overhead at the RB 355). For example, SP 305 may manage ordering of
waves (e.g., SP 305 may manage allocation and reallocation of wave
slots for processing waves, based on terminated waves) at the wave
scheduler 310, and SP 305 may output processed results associated
with the waves (e.g., SP 305 may output pixels for non-terminated
waves, SP 305 may output a number of invocations field for
terminated waves, etc.) to the RB 355 via the distributor 325.
[0081] In some cases, GPU 300 (e.g., RB 355) may process the waves
(e.g., manage rendering associated with the waves, process pixels
of the waves) based on the order in which the waves are output to
the RB 355 by the distributor 325. In an example aspect, processing
pixels of Wave W4 may include performing one or more rendering
passes for rendering pixels of Wave W4. In some aspects, GPU 300
(e.g., RB 355) may process the waves based on terminated wave bits
and the number of invocations field received from the SP 305 (e.g.,
as received from the distributor 325). For example, based on a
terminated wave bit and the number of invocations field, the RB 355
may identify a terminated wave and the number of pixels associated
with the terminated wave (e.g., such that slot resources associated
with the terminated wave may be reallocated while maintaining
ordering information at RB 355).
[0082] FIG. 4A illustrates an example of a processing diagram 400
illustrating some aspects of out of order wave slot release for a
terminated wave, in accordance with aspects of the present
disclosure. In some examples, processing diagram 400 may implement
aspects of multimedia system 100. Processing diagram 400 may
implement aspects of devices 105, device 200, and GPU 300 described
herein. Processing diagram 400 includes a sequence queue 405, wave
slots 415, and an execution timeline 420. Some example aspects of
processing diagram 400 are also described in reference to FIG.
3.
[0083] Sequence queue 405 may include a read pointer (RPTR) 406, a
write pointer (WPTR) 407, and entries 410 (e.g., queue
assignments). RPTR 406 may include a pointer associated with
reading data (e.g., wave information, assigned slot) from the
sequence queue 405. WPTR 407 may include a pointer associated with
writing data to the sequence queue 405. Sequence queue 405, for
example, may include a lookup table, a location in local memory,
etc., and RPTR 406 and WPTR 407 may include pointers associated
with reading data from the sequence queue 405 and writing data to
the sequence queue 405.
[0084] In some aspects, sequence queue 405 may be determined or
modified by a GPU (e.g., GPU 225, GPU 300) as described herein. For
example, sequence queue 405 may be determined or modified by an SP
(e.g., SP 305) as described herein. For example, a wave scheduler
(e.g., wave scheduler 310) as described herein may determine or
modify sequence queue 405.
[0085] Entries 410 may each indicate a wave scheduled for
processing (e.g., as scheduled by wave scheduler 310 of SP 305), a
slot allocated for processing the wave, or a combination thereof.
In the example illustrated in processing diagram 400, entries 410
indicated in the sequence queue 405 may be associated with waves
(e.g., waves W0 through W6, which may also be referred to as
Pixel_warp W0 through Pixel_warp W6) scheduled to be processed and
slots (e.g., slots S0 through S3) which may be allocated for
processing the waves. For example, entry 410-a (e.g., W0/S0) may be
associated with wave W0 and a slot S0 allocated for processing the
wave W0, entry 410-b (e.g., W1/S1) may be associated with a wave W1
and a slot S1 allocated for processing the wave W1, entry 410-c
(e.g., W2/S2) may be associated with a wave W2 and a slot S2
allocated for processing the wave W2, etc.
[0086] Slots 415 may correspond to wave slots which may be
allocated by SP 305 for processing input waves. For example, as
illustrated in the example of FIG. 4A, slot 415-a may correspond to
slot S0, slot 415-b may correspond to slot S1, slot 415-c may
correspond to slot S2, and slot 415-d may correspond to slot
S3.
[0087] Execution timeline 420 is an example timeline illustrating
aspects of processing waves W0 through W6 by GPU 300 (e.g., SP
305). Timelines 425-a through 425-g correspond to timelines
associated with processing waves W0 through W6. For example,
timeline 425-a may be associated with processing wave W0
(Pixel_warp W0), timeline 425-b may be associated with processing
wave W1 (Pixel_warp W1), timeline 425-c may be associated with
processing wave W2 (Pixel_warp W2), etc.
[0088] In the example illustrated in FIG. 4A, waves W1 through W3
have a highest priority (e.g., according to input order) among
waves W0 through W6 to be processed. Waves W1 through W3 are
processed at slots 415-a through 415-d (e.g., slots S0 through S3),
respectively. Waves W4 through W6 are processed at slots 415-a
through 415-c (e.g., slots S0 through S2), respectively. In other
words, processing diagram 400 may illustrate a wave priority
corresponding to an eldest scheme (e.g., where older waves have
priority over newer waves).
[0089] In the example, a next wave (e.g., wave W4 having the next
highest priority following waves W0 through W3) is allocated slot
415-a (e.g., slot S0) for processing, after processing of wave W0
(e.g., illustrated by timeline 425-a), which has the highest
priority, is completed.
[0090] Wave W1 (Pixel_warp W1) being processed in slot 415-a (e.g.,
wave slot S1) is terminated (as illustrated by timeline 425-b)
prior to processing completion of wave W0. However, slot 415-a is
not reallocated for processing another wave to be processed, until
after a wait period 430. That is, GPU 300 waits to allocate a next
wave (e.g., wave W5 having the next highest priority following
waves W0 through W4) to slot 415-b (e.g., slot S1) for processing
until after the wait period 430 has expired. Specifically, the slot
415-b may be held idle until older waves have been executed (e.g.,
until wave W0 has been executed, completed, etc.). For example, in
order to output color information associated with waves W0 through
W6, without the techniques described herein, slot 415-b, upon
termination of wave W1, may remain idle. Accordingly, slot 415-b
(e.g., slot S1) may incur an idle time corresponding to the wait
period 430. As illustrated in FIG. 4A, without out of order wave
slot release for a terminated wave, an early executing wave (e.g.,
a wave having relatively high priority among waves to be processed)
may be the last to be processed completely among waves to be
processed, which may result in low resource utilization of an SP
(e.g., low resource utilization of SP 305).
[0091] FIG. 4B illustrates an example of a processing diagram 401
that supports out of order wave slot release for a terminated wave
in accordance with aspects of the present disclosure. In some
examples, processing diagram 401 may implement aspects of
multimedia system 100. Processing diagram 401 may implement aspects
of devices 105, device 200, and GPU 300 described herein. Example
aspects of processing diagram 401 are described in reference to
FIG. 3.
[0092] Processing diagram 401 illustrates an example of a sequence
queue 435, wave slots 445, and an execution timeline 450. Example
aspects of processing diagram 401 are described in reference to
FIG. 3. Sequence queue 435 may include a RPTR 436, a WPTR 407, and
entries 440. Aspects of sequence queue 435, RPTR 436, WPTR 407,
entries 440, slots 445, execution timeline 450, and timelines 455
may include aspects similar to those described with respect to
sequence queue 405, RPTR 406, WPTR 407, entries 410, slots 415,
execution timeline 420, and timelines 425, but may differ where
indicated in relation to aspects as described herein.
[0093] In some example aspects, sequence queue 435 may be
determined or modified by a device 105, for example, a GPU (e.g.,
GPU 225, GPU 300) as described herein. For example, a wave
scheduler (e.g., wave scheduler 310) as described herein may modify
sequence queue 435 based on a terminated wave (e.g., based on
termination of a wave W1 allocated a slot S1). For example, in some
cases, wave scheduler 310 may add or toggle a terminated wave bit
to sequence queue 435, may update sequence queue 435 based on out
of order wave slot release, etc.
[0094] Referring to the example of FIG. 4B, the device 105 (e.g.,
GPU 300) may perform a test operation for a wave W1 (e.g., a first
wave). The test operation may include, for example, a visibility
pass associated with primitives that visible in an image and
primitives that are invisible in the same image. In some aspects,
GPU 300 may identify a number of pixels associated with the wave W1
based on the test operation. Based on the test operation, GPU 300
may determine to terminate all pixels of the identified number of
pixels (e.g., the test operation may result in GPU 300
determination that all pixels associated with the wave are
invisible, can be terminated, etc.). As such, in addition to
setting the wave terminated bit, the number of pixels may be
indicated, for example, by a quad count number (e.g., a number of
invocations field) associated with the wave W1. In some example
aspects, GPU 300 may perform multiple test operations.
[0095] In some example aspects, the device 105 (e.g., GPU 300) may
determine, based on the test operation, to terminate wave W1. Wave
W1 may be associated with a slot S1 (e.g., slot 445-b). In some
aspects, the determination to terminate wave W1 may be based on the
determination to terminate all pixels of the identified number of
pixels.
[0096] The device 105 (e.g., a SP 305 of GPU 300) may identify a
wave order associated with a set of waves including wave W1 based
on sequence queue 435. For example, various aspects or components
of GPU 300 (e.g., a SP, a distributor, a dispatcher, a RB) may
identify a wave order associated with waves W1 through W6 based on
sequence queue 435. In some aspects, GPU 300 components may
identify the terminated wave bit based on the identified wave order
of sequence queue 435. For example, in some cases, a distributor or
a dispatcher of a GPU may output information based on sequence
queue 435. Based on sequence queue 435, the distributor may output
pixel information (e.g., color information) of W0 to a RB, may
identify a killed wave bit (terminated wave bit) and output a
number of invocations field to the RB, may output pixel information
(e.g., color information) of W2 to the RB, etc. As such, slot 445-b
may be released (e.g., for processing of W4 with reduced latency or
idle time) while maintaining output order to a downstream block
(e.g., such as a RB). The RB may thus perform post color operations
efficiently, as the RB may receive information in order and
identify any terminated waves/pixels such that pixel ordering may
be maintained (e.g., such that pixel information or color
information associated with wave W0 and wave W2 may be
appropriately handled/ordered by the RB based on the number of
invocations/pixels field associated with wave W1).
[0097] The device 105 (e.g., GPU 300) may update a terminated wave
bit (e.g., "killed", as indicated at entry 440-b) associated with
slot S1 based on the determination to terminate wave W1. For
example, in updating the terminated wave bit, GPU 300 may write the
terminated wave bit and the number of invocations field to entry
440-b of sequence queue 435 corresponding to wave W1. In some
aspects, the device 105 (e.g., GPU 300) may update a number of
invocations field associated with wave W1 based on the
determination to terminate wave W1. The number of invocations field
may include a quad count number, a pixel count number, a fragments
number, etc. for indicating the number of killed invocations,
pixels, quads, etc. In some examples, the device 105 (e.g., GPU
300) may update the number of invocations field associated with
wave W1 to indicate the identified number of pixels associated with
the wave W1 (e.g., the identified pixels to be terminated).
[0098] Generally, a SP may determine the number of invocations
(e.g., pixels) associated with the terminated wave (e.g., via a
test operation, etc.), and any component may store the number of
invocations field (e.g., not necessarily the sequence queue 435).
For example, in some cases, the sequence queue 435 may include the
terminated wave bit, such that a distributor or dispatcher, upon
identifying the terminated wave bit, may output the number of
invocations field (e.g., instead of any color information
associated with the terminated wave). In some cases, a distributor
or dispatcher of the SP may gather or determine color information
(e.g., for waves W0, W2, W3, etc.) and number of invocations field
(e.g., for wave W1) from other components of the SP, and the
distributor or dispatcher may dispatch such information to a RB
based on the sequence queue 435 (e.g., based on an ordering
indicated by sequence queue 435, based on any wave terminated bits
of sequence queue 435, etc.).
[0099] According to example aspects of out of order wave slot
release for a terminated wave (e.g., terminated wave W1) as
described herein, the device 105 (e.g., GPU 300, wave scheduler
310) may release slot S1, for example, based on updating the
terminated wave bit and the number of invocations field. For
example, the device 105 (e.g., GPU 300, wave scheduler 310) may
release slot S1 for incoming workloads (e.g., other waves to be
processed).
[0100] In some aspects, the device 105 (e.g., SP 305) may process
one or more pixels of wave W2 (e.g., a second wave) using slot S2
(e.g., slot 445-c). In some example aspects, GPU 300 may process
one or more pixels of wave W5 (e.g., a third wave) using the
released slot S1 (e.g., slot 445-b). In an example, GPU 300 may
process one or more pixels of wave W4 (e.g., a fourth wave) using
slot S3.
[0101] In some example aspects, the device 105 (e.g., GPU 300) may
write the terminated wave bit (e.g., "killed") and the number of
invocations field to a first entry 440-b of sequence queue 435,
where first entry 440-b corresponds to the wave W1 (i.e., the
terminated wave). The device 105 (e.g., GPU 300) may write one or
more pixels of the wave W2 (e.g., the second wave) to a second
entry 440-c of sequence queue 435, where second entry 440-c
corresponds to wave W2. The one or more pixels may include, for
example, color information as described herein.
[0102] In some aspects, the device 105 (e.g., GPU 300) may write
the one or more pixels of the wave W5 (e.g., the third wave) to a
third entry 440-f of sequence queue 435, where third entry 440-e
corresponds to the wave W5. The device 105 (e.g., GPU 300) may
write the one or more pixels of wave W4 (e.g., the fourth wave) to
a fourth entry 440-e of the sequence queue 435, where fourth entry
440-e corresponds to the wave W4.
[0103] The device 105 (e.g., GPU 300) may output, by SP 305 (e.g.,
by distributor 325), the entries 440 to the rendering backend
(e.g., RB 355) of the device based on the sequence queue 435. In
some aspects, the device 105 (e.g., GPU 300, RB 355) may identify a
pixel ordering for rendering pixel information based on the number
of invocations field associated with the wave W1 (e.g., the first
wave). In some example aspects, the device 105 (e.g., GPU 300) may
output the number of invocations field to a rendering backend
(e.g., RB 355) of the device 105, based on the terminated wave bit.
In some aspects, the device 105 (e.g., GPU 300) may output the
number of invocations field based on the identified terminated wave
bit and the identified wave order. The device 105 (e.g., GPU 300)
may receive, at the rendering backend (e.g., RB 355), the output
number of invocations field. The device 105 (e.g., GPU 300, RB 355)
may identify a pixel ordering for rendering based on the received
number of invocations field.
[0104] In some example aspects, for example, the device 105 (e.g.,
GPU 300) may perform test operations as described herein, for waves
other than wave W1 (e.g., for a wave W2). For example, the device
105 (e.g., GPU 300) may perform a test operation as described
herein, for wave W2 (e.g., the second wave), and may identify a
number of pixels associated with the wave W2 based on the second
test operation. The device 105 (e.g., GPU 300) may process a subset
of pixels associated with the wave W2, for example, where
processing includes outputting color information associated with
the subset of pixels to the rendering backend (e.g., RB 355) of the
device.
[0105] In some aspects, there may be multiple scenarios associated
with enabling a pixel kill feature for processing waves. One
scenario may include a partial wave kill (e.g., some valid pixels
survive). In such a scenario, techniques may include optimization
to reassemble survived pixels into full waves, which may increase
wave execution efficiency but at the cost of additional overhead
due to "multiple pass" execution (e.g., stopping a wave execution
immediately after the wave is killed, pushing surviving pixels to a
survivor pool, reassembling pixels from the survivor pool). Another
example scenario may include a complete wave kill (e.g., no valid
pixel survives). In such a scenario, where all valid quads in the
wave are killed, the techniques described herein may be applied to
pass down minimal information (e.g., a number of invocations field,
a terminated wave bit, etc.) to a downstream block (e.g., RB 355).
For example, the techniques include providing bits associated with
a pixel quad counter, which may be relatively minimal compared to
data (e.g., pixel information, color information, etc.) associated
with a wave as a whole.
[0106] FIG. 5 shows a block diagram 500 of a device 505 that
supports out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure. The device 505
may be an example of aspects of a device (e.g., a device 105 or
device 200) as described herein. The device 505 may include a CPU
510, a GPU 515, and a display 520. The device 505 may also include
a processor. Each of these components may be in communication with
one another (e.g., via one or more buses).
[0107] CPU 510 may be an example of CPU 210 described with
reference to FIG. 2. CPU 510 may execute one or more software
applications, such as web browsers, graphical user interfaces,
video games, or other applications involving graphics rendering for
image depiction (e.g., via display 520). As described herein, CPU
510 may encounter a GPU program (e.g., a program suited for
handling by GPU 515) when executing the one or more software
applications. Accordingly, CPU 510 may submit rendering commands to
GPU 515 (e.g., via a GPU driver containing a compiler for parsing
API-based commands).
[0108] The GPU 515 may determine, based on a test operation, to
terminate a first wave associated with a first slot of a set of
slots, update a terminated wave bit associated with the first slot
based on the determination to terminate the first wave, update a
number of invocations field associated with the first wave based on
the determination to terminate the first wave, release the first
slot based on updating the terminated wave bit and the number of
invocations field, and output the number of invocations field to a
rendering backend of the device based on the terminated wave bit.
The GPU 515 may be an example of aspects of the GPU 810 described
herein.
[0109] The GPU 515, or its sub-components, may be implemented in
hardware, code (e.g., software or firmware) executed by a
processor, or any combination thereof. If implemented in code
executed by a processor, the functions of the GPU 515, or its
sub-components may be executed by a general-purpose processor, a
DSP, an ASIC, a FPGA or other programmable logic device, discrete
gate or transistor logic, discrete hardware components, or any
combination thereof designed to perform the functions described in
the present disclosure.
[0110] The GPU 515, or its sub-components, may be physically
located at various positions, including being distributed such that
portions of functions are implemented at different physical
locations by one or more physical components. In some examples, the
GPU 515, or its sub-components, may be a separate and distinct
component in accordance with various aspects of the present
disclosure. In some examples, the GPU 515, or its sub-components,
may be combined with one or more other hardware components,
including but not limited to an input/output (I/O) component, a
transceiver, a network server, another computing device, one or
more other components described in the present disclosure, or a
combination thereof in accordance with various aspects of the
present disclosure.
[0111] Display 520 may display content generated by other
components of the device. Display 520 may be an example of display
245 as described with reference to FIG. 2. In some examples,
display 520 may be connected with a display buffer which stores
rendered data until an image is ready to be displayed (e.g., as
described with reference to FIG. 2). The display 520 may illuminate
according to signals or information generated by other components
of the device 505. For example, the display 520 may receive display
information (e.g., pixel mappings, display adjustments) from GPU
515, and may illuminate accordingly. The display 520 may represent
a unit capable of displaying video, images, text or any other type
of data for consumption by a viewer. Display 520 may include a
liquid-crystal display (LCD), a light emitting diode (LED) display,
an organic LED (OLED), an active-matrix OLED (AMOLED), or the like.
In some cases, display 520 and an I/O controller (e.g., I/O
controller 815) may be or represent aspects of a same component
(e.g., a touchscreen) of device 505.
[0112] FIG. 6 shows a block diagram 600 of a device 605 that
supports out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure. The device 605
may be an example of aspects of a device 505, a device 105, or a
device 200 as described herein. The device 605 may include a CPU
610, a GPU 615, and a display 635. The device 605 may also include
a processor. Each of these components may be in communication with
one another (e.g., via one or more buses).
[0113] CPU 610 may be an example of CPU 210 described with
reference to FIG. 2. CPU 610 may execute one or more software
applications, such as web browsers, graphical user interfaces,
video games, or other applications involving graphics rendering for
image depiction (e.g., via display 635). As described herein, CPU
610 may encounter a GPU program (e.g., a program suited for
handling by GPU 615) when executing the one or more software
applications. Accordingly, CPU 610 may submit rendering commands to
GPU 615 (e.g., via a GPU driver containing a compiler for parsing
API-based commands).
[0114] The GPU 615 may be an example of aspects of the GPU 515 as
described herein. The GPU 615 may include a test operation manager
620, a sequencer 625, and a streaming processor 630. The GPU 615
may be an example of aspects of the GPU 810 described herein.
[0115] The test operation manager 620 may determine, based on a
test operation, to terminate a first wave associated with a first
slot of a set of slots. The sequencer 625 may update a terminated
wave bit associated with the first slot based on the determination
to terminate the first wave and update a number of invocations
field associated with the first wave based on the determination to
terminate the first wave. The streaming processor 630 may release
the first slot based on updating the terminated wave bit and the
number of invocations field and output the number of invocations
field to a rendering backend of the device based on the terminated
wave bit.
[0116] The display 635 may transmit signals generated by other
components of the device 605. In some examples, the display 635 may
be collocated with a CPU 610 in a transceiver module. For example,
the display 635 may be an example of aspects of the transceiver 820
described with reference to FIG. 8. The display 635 may utilize a
single antenna or a set of antennas.
[0117] Display 635 may display content generated by other
components of the device. Display 635 may be an example of display
245 as described with reference to FIG. 2. In some examples,
display 635 may be connected with a display buffer which stores
rendered data until an image is ready to be displayed (e.g., as
described with reference to FIG. 2). The display 635 may illuminate
according to signals or information generated by other components
of the device 605. For example, the display 635 may receive display
information (e.g., pixel mappings, display adjustments) from GPU
615, and may illuminate accordingly. The display 635 may represent
a unit capable of displaying video, images, text or any other type
of data for consumption by a viewer. Display 635 may include a
liquid-crystal display (LCD), a light emitting diode (LED) display,
an organic LED (OLED), an active-matrix OLED (AMOLED), or the like.
In some cases, display 635 and an I/O controller (e.g., I/O
controller 815) may be or represent aspects of a same component
(e.g., a touchscreen) of device 605.
[0118] FIG. 7 shows a block diagram 700 of a GPU 705 that supports
out of order wave slot release for a terminated wave in accordance
with aspects of the present disclosure. The GPU 705 may be an
example of aspects of a GPU 515, a GPU 615, or a GPU 810 described
herein. The GPU 705 may include a test operation manager 710, a
sequencer 715, a streaming processor 720, a dispatcher 725, and a
rendering backend 730. Each of these modules may communicate,
directly or indirectly, with one another (e.g., via one or more
buses).
[0119] The test operation manager 710 may determine, based on a
test operation, to terminate a first wave associated with a first
slot of a set of slots. In some examples, the test operation
manager 710 may perform the test operation for the first wave. In
some examples, the test operation manager 710 may identify a number
of pixels associated with the first wave based on the test
operation.
[0120] In some examples, the test operation manager 710 may
determine, based on the test operation, to terminate all pixels of
the identified number of pixels, where the determination to
terminate the first wave is based on the determination to terminate
all pixels of the identified number of pixels. In some examples,
the test operation manager 710 may perform a second test operation
for the second wave. In some examples, the test operation manager
710 may identify a number of pixels associated with the second wave
based on the second test operation, where the one or more pixels
include a subset of the number of pixels associated with the second
wave.
[0121] The sequencer 715 may update a terminated wave bit
associated with the first slot based on the determination to
terminate the first wave. In some examples, the sequencer 715 may
update a number of invocations field associated with the first wave
based on the determination to terminate the first wave. In some
examples, the sequencer 715 may update the number of invocations
field associated with the first wave to indicate the identified
number of pixels. In some examples, the sequencer 715 may write the
terminated wave bit and the number of invocations field to an entry
of a sequence queue corresponding to the first wave. In some
examples, the sequencer 715 may write the terminated wave bit and
the number of invocations field to a first entry of a sequence
queue corresponding to the first wave.
[0122] In some examples, the sequencer 715 may write the one or
more pixels of the second wave to a second entry of the sequence
queue corresponding to the second wave. In some examples, the
sequencer 715 may write the one or more pixels of the third wave to
a third entry of the sequence queue corresponding to the third
wave. In some examples, the sequencer 715 may write the one or more
pixels of the fourth wave to a fourth entry of the sequence queue
corresponding to the fourth wave.
[0123] The streaming processor 720 may release the first slot based
on updating the terminated wave bit and the number of invocations
field. In some examples, the streaming processor 720 may output the
number of invocations field to a rendering backend of the device
based on the terminated wave bit. In some examples, the streaming
processor 720 may process one or more pixels of a second wave using
a second slot. In some examples, the streaming processor 720 may
process one or more pixels of a third wave using the released first
slot. In some examples, the streaming processor 720 may process one
or more pixels of a fourth wave using a third slot. In some
examples, the streaming processor 720 may output, by a streaming
processor of the device, the first entry, the second entry, the
third entry, and the fourth entry to the rendering backend of the
device based on the sequence queue.
[0124] The dispatcher 725 may identify a wave order associated with
a set of waves based on a sequence queue, where the set of waves
includes at least the first wave. In some examples, the dispatcher
725 may identify the terminated wave bit based on the identified
wave order, where the number of invocations field is output based
on the identified terminated wave bit and the identified wave
order. In some examples, the dispatcher 725 may identify the
terminated wave bit based on the sequence queue, where the number
of invocations field is output based on the identified terminated
wave bit and the identified wave order. In some examples, the
dispatcher 725 may output color information associated with the one
or more pixels to the rendering backend of the device based on
processing the one or more pixels of the second wave.
[0125] The rendering backend 730 may receive the output number of
invocations field. In some examples, the rendering backend 730 may
identify a pixel ordering for rendering based on the received
number of invocations field. In some examples, the rendering
backend 730 may identify a pixel ordering for rendering pixel
information based on the number of invocations field associated
with the first wave.
[0126] FIG. 8 shows a diagram of a system 800 including a device
805 that supports out of order wave slot release for a terminated
wave in accordance with aspects of the present disclosure. The
device 805 may be an example of or include the components of device
505, device 605, or a device as described herein. The device 805
may include components for bi-directional voice and data
communications including components for transmitting and receiving
communications, including a GPU 810, an I/O controller 815, memory
830, and a processor or CPU 840. In some cases, device 805 may
include a transceiver 820. These components may be in electronic
communication via one or more buses (e.g., bus 845).
[0127] The GPU 810 may determine, based on a test operation, to
terminate a first wave associated with a first slot of a set of
slots, update a terminated wave bit associated with the first slot
based on the determination to terminate the first wave, update a
number of invocations field associated with the first wave based on
the determination to terminate the first wave, release the first
slot based on updating the terminated wave bit and the number of
invocations field, and output the number of invocations field to a
rendering backend of the device based on the terminated wave
bit.
[0128] CPU 840 may include an intelligent hardware device, (e.g., a
general-purpose processor, a DSP, a microcontroller, an ASIC, an
FPGA, a programmable logic device, a discrete gate or transistor
logic component, a discrete hardware component, or any combination
thereof). In some cases, CPU 840 may be configured to operate a
memory array using a memory controller. In other cases, a memory
controller may be integrated into CPU 840. CPU 840 may be
configured to execute computer-readable instructions stored in a
memory to perform various functions (e.g., functions or tasks
supporting dynamic bin ordering for load synchronization).
[0129] The I/O controller 815 may manage input and output signals
for the device 805. The I/O controller 815 may also manage
peripherals not integrated into the device 805. In some cases, the
I/O controller 815 may represent a physical connection or port to
an external peripheral. In some cases, the I/O controller 815 may
utilize an operating system such as iOS.RTM., ANDROID.RTM.,
MS-DOS.RTM., MS-WINDOWS.RTM., OS/2.RTM., UNIX.RTM., LINUX.RTM., or
another known operating system. In other cases, the I/O controller
815 may represent or interact with a modem, a keyboard, a mouse, a
touchscreen, or a similar device. In some cases, the I/O controller
815 may be implemented as part of a processor. In some cases, a
user may interact with the device 805 via the I/O controller 815 or
via hardware components controlled by the I/O controller 815.
[0130] The transceiver 820 may communicate bi-directionally, via
one or more antennas, wired, or wireless links as described herein.
For example, the transceiver 820 may represent a wireless
transceiver and may communicate bi-directionally with another
wireless transceiver. The transceiver 820 may also include a modem
to modulate the packets and provide the modulated packets to the
antennas for transmission, and to demodulate packets received from
the antennas.
[0131] The memory 830 may include RAM and ROM. The memory 830 may
store computer-readable, computer-executable code or software 835
including instructions that, when executed, cause the processor to
perform various functions described herein. In some cases, the
memory 830 may contain, among other things, a BIOS which may
control basic hardware or software operation such as the
interaction with peripheral components or devices.
[0132] In some cases, the GPU 810 and/or the CPU 840 may include an
intelligent hardware device, (e.g., a general-purpose processor, a
DSP, a microcontroller, an ASIC, an FPGA, a programmable logic
device, a discrete gate or transistor logic component, a discrete
hardware component, or any combination thereof). In some cases, the
GPU 810 and/or the CPU 840 may be configured to operate a memory
array using a memory controller. In other cases, a memory
controller may be integrated into the GPU 810 and/or the CPU 840.
The GPU 810 and/or the CPU 840 may be configured to execute
computer-readable instructions stored in a memory (e.g., the memory
830) to cause the device 805 to perform various functions (e.g.,
functions or tasks supporting out of order wave slot release for a
terminated wave).
[0133] The software 835 may include instructions to implement
aspects of the present disclosure, including instructions to
support image processing. The software 835 may be stored in a
non-transitory computer-readable medium such as system memory or
other type of memory. In some cases, the software 835 may not be
directly executable by the CPU 840 but may cause a computer (e.g.,
when compiled and executed) to perform functions described
herein.
[0134] FIG. 9 shows a flowchart illustrating a method 900 that
supports out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure. The operations
of method 900 may be implemented by a device or its components as
described herein. For example, the operations of method 900 may be
performed by a GPU as described with reference to FIGS. 5 through
8. In some examples, a device may execute a set of instructions to
control the functional elements of the device to perform the
functions described herein. Additionally or alternatively, a device
may perform aspects of the functions described herein using
special-purpose hardware.
[0135] At 905, the device may determine, based on a test operation,
to terminate a first wave associated with a first slot of a set of
slots. The operations of 905 may be performed according to the
methods described herein. In some examples, aspects of the
operations of 905 may be performed by a test operation manager as
described with reference to FIGS. 5 through 8.
[0136] At 910, the device may update a terminated wave bit
associated with the first slot based on the determination to
terminate the first wave. The operations of 910 may be performed
according to the methods described herein. In some examples,
aspects of the operations of 910 may be performed by a sequencer as
described with reference to FIGS. 5 through 8.
[0137] At 915, the device may update a number of invocations field
associated with the first wave based on the determination to
terminate the first wave. The operations of 915 may be performed
according to the methods described herein. In some examples,
aspects of the operations of 915 may be performed by a sequencer as
described with reference to FIGS. 5 through 8.
[0138] At 920, the device may release the first slot based on
updating the terminated wave bit and the number of invocations
field. The operations of 920 may be performed according to the
methods described herein. In some examples, aspects of the
operations of 920 may be performed by a streaming processor as
described with reference to FIGS. 5 through 8.
[0139] At 925, the device may output the number of invocations
field to a rendering backend of the device based on the terminated
wave bit. The operations of 925 may be performed according to the
methods described herein. In some examples, aspects of the
operations of 925 may be performed by a streaming processor as
described with reference to FIGS. 5 through 8.
[0140] FIG. 10 shows a flowchart illustrating a method 1000 that
supports out of order wave slot release for a terminated wave in
accordance with aspects of the present disclosure. The operations
of method 1000 may be implemented by a device or its components as
described herein. For example, the operations of method 1000 may be
performed by a GPU as described with reference to FIGS. 5 through
8. In some examples, a device may execute a set of instructions to
control the functional elements of the device to perform the
functions described herein. Additionally or alternatively, a device
may perform aspects of the functions described herein using
special-purpose hardware.
[0141] At 1005, the device may determine, based on a test
operation, to terminate a first wave associated with a first slot
of a set of slots. The operations of 1005 may be performed
according to the methods described herein. In some examples,
aspects of the operations of 1005 may be performed by a test
operation manager as described with reference to FIGS. 5 through
8.
[0142] At 1010, the device may update a terminated wave bit
associated with the first slot based on the determination to
terminate the first wave. The operations of 1010 may be performed
according to the methods described herein. In some examples,
aspects of the operations of 1010 may be performed by a sequencer
as described with reference to FIGS. 5 through 8.
[0143] At 1015, the device may update a number of invocations field
associated with the first wave based on the determination to
terminate the first wave. The operations of 1015 may be performed
according to the methods described herein. In some examples,
aspects of the operations of 1015 may be performed by a sequencer
as described with reference to FIGS. 5 through 8.
[0144] At 1020, the device may release the first slot based on
updating the terminated wave bit and the number of invocations
field. The operations of 1020 may be performed according to the
methods described herein. In some examples, aspects of the
operations of 1020 may be performed by a streaming processor as
described with reference to FIGS. 5 through 8.
[0145] At 1025, the device (e.g., a dispatcher of the device) may
identify a wave order associated with a set of waves based on a
sequence queue, where the set of waves includes at least the first
wave. The operations of 1025 may be performed according to the
methods described herein. In some examples, aspects of the
operations of 1025 may be performed by a dispatcher as described
with reference to FIGS. 5 through 8.
[0146] At 1030, the device (e.g., a dispatcher of the device) may
identify the terminated wave bit based on the identified wave
order, where the number of invocations field is output based on the
identified terminated wave bit and the identified wave order. The
operations of 1030 may be performed according to the methods
described herein. In some examples, aspects of the operations of
1030 may be performed by a dispatcher as described with reference
to FIGS. 5 through 8.
[0147] At 1035, the device may output the number of invocations
field to a rendering backend of the device based on the identified
terminated wave bit and the identified wave order. The operations
of 1035 may be performed according to the methods described herein.
In some examples, aspects of the operations of 1035 may be
performed by a streaming processor as described with reference to
FIGS. 5 through 8.
[0148] At 1040, the device (e.g., a rendering backend of the
device) may receive the output number of invocations field. The
operations of 1040 may be performed according to the methods
described herein. In some examples, aspects of the operations of
1040 may be performed by a rendering backend as described with
reference to FIGS. 5 through 8.
[0149] At 1045, the device (e.g., a rendering backend of the
device) may identify a pixel ordering for rendering (e.g., by the
rendering backend) based on the received number of invocations
field. The operations of 1045 may be performed according to the
methods described herein. In some examples, aspects of the
operations of 1045 may be performed by a rendering backend as
described with reference to FIGS. 5 through 8.
[0150] It should be noted that the methods described herein
describe possible implementations, and that the operations and the
steps may be rearranged or otherwise modified and that other
implementations are possible. Further, aspects from two or more of
the methods may be combined.
[0151] Information and signals described herein may be represented
using any of a variety of different technologies and techniques.
For example, data, instructions, commands, information, signals,
bits, symbols, and chips that may be referenced throughout the
description may be represented by voltages, currents,
electromagnetic waves, magnetic fields or particles, optical fields
or particles, or any combination thereof.
[0152] The various illustrative blocks and modules described in
connection with the disclosure herein may be implemented or
performed with a general-purpose processor, a DSP, an ASIC, an
FPGA, or other programmable logic device, discrete gate or
transistor logic, discrete hardware components, or any combination
thereof designed to perform the functions described herein. A
general-purpose processor may be a microprocessor, but in the
alternative, the processor may be any conventional processor,
controller, microcontroller, or state machine. A processor may also
be implemented as a combination of computing devices (e.g., a
combination of a DSP and a microprocessor, multiple
microprocessors, one or more microprocessors in conjunction with a
DSP core, or any other such configuration).
[0153] The functions described herein may be implemented in
hardware, software executed by a processor, firmware, or any
combination thereof. If implemented in software executed by a
processor, the functions may be stored on or transmitted over as
one or more instructions or code on a computer-readable medium.
Other examples and implementations are within the scope of the
disclosure and appended claims. For example, due to the nature of
software, functions described herein can be implemented using
software executed by a processor, hardware, firmware, hardwiring,
or combinations of any of these. Features implementing functions
may also be physically located at various positions, including
being distributed such that portions of functions are implemented
at different physical locations.
[0154] Computer-readable media includes both non-transitory
computer storage media and communication media including any medium
that facilitates transfer of a computer program from one place to
another. A non-transitory storage medium may be any available
medium that can be accessed by a general purpose or special purpose
computer. By way of example, and not limitation, non-transitory
computer-readable media may include RAM, ROM, electrically erasable
programmable ROM (EEPROM), flash memory, compact disk (CD) ROM or
other optical disk storage, magnetic disk storage or other magnetic
storage devices, or any other non-transitory medium that can be
used to carry or store desired program code means in the form of
instructions or data structures and that can be accessed by a
general-purpose or special-purpose computer, or a general-purpose
or special-purpose processor. Also, any connection is properly
termed a computer-readable medium. For example, if the software is
transmitted from a website, server, or other remote source using a
coaxial cable, fiber optic cable, twisted pair, digital subscriber
line (DSL), or wireless technologies such as infrared, radio, and
microwave, then the coaxial cable, fiber optic cable, twisted pair,
DSL, or wireless technologies such as infrared, radio, and
microwave are included in the definition of medium. Disk and disc,
as used herein, include CD, laser disc, optical disc, digital
versatile disc (DVD), floppy disk and Blu-ray disc where disks
usually reproduce data magnetically, while discs reproduce data
optically with lasers. Combinations of the above are also included
within the scope of computer-readable media.
[0155] As used herein, including in the claims, "or" as used in a
list of items (e.g., a list of items prefaced by a phrase such as
"at least one of" or "one or more of") indicates an inclusive list
such that, for example, a list of at least one of A, B, or C means
A or B or C or AB or AC or BC or ABC (i.e., A and B and C). Also,
as used herein, the phrase "based on" shall not be construed as a
reference to a closed set of conditions. For example, an exemplary
step that is described as "based on condition A" may be based on
both a condition A and a condition B without departing from the
scope of the present disclosure. In other words, as used herein,
the phrase "based on" shall be construed in the same manner as the
phrase "based at least in part on."
[0156] In the appended figures, similar components or features may
have the same reference label. Further, various components of the
same type may be distinguished by following the reference label by
a dash and a second label that distinguishes among the similar
components. If just the first reference label is used in the
specification, the description is applicable to any one of the
similar components having the same first reference label
irrespective of the second reference label, or other subsequent
reference label.
[0157] The description set forth herein, in connection with the
appended drawings, describes example configurations and does not
represent all the examples that may be implemented or that are
within the scope of the claims. The term "exemplary" used herein
means "serving as an example, instance, or illustration," and not
"preferred" or "advantageous over other examples." The detailed
description includes specific details for the purpose of providing
an understanding of the described techniques. These techniques,
however, may be practiced without these specific details. In some
instances, well-known structures and devices are shown in block
diagram form in order to avoid obscuring the concepts of the
described examples.
[0158] The description herein is provided to enable a person
skilled in the art to make or use the disclosure. Various
modifications to the disclosure will be readily apparent to those
skilled in the art, and the generic principles defined herein may
be applied to other variations without departing from the scope of
the disclosure. Thus, the disclosure is not limited to the examples
and designs described herein, but is to be accorded the broadest
scope consistent with the principles and novel features disclosed
herein.
* * * * *